1 / 46

Study of CP Violation at BABAR

Study of CP Violation at BABAR. David Lange Lawrence Livermore National Laboratory For the BABAR Collaboration. 30th SLAC Summer Institute, August 5-16,2002. Coupling for Q W + q is ~ V * Qq SM mechanism for CP violation is non-0 complex phase in V CKM .

deiter
Download Presentation

Study of CP Violation at BABAR

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Study of CP Violation at BABAR David Lange Lawrence Livermore National Laboratory For the BABAR Collaboration 30th SLAC Summer Institute, August 5-16,2002

  2. Coupling for Q W+ q is ~ V*Qq SM mechanism for CP violation is non-0 complex phase in VCKM. Orthogonality of the 1st and 3rd columns: B0J/yKS * VtdVtb * VudVub * VcdVcb CKM matrix and Unitarity Triangle a(f2) B0pp, rp Channels that can be used to measure a,b,g B0DK g(f3) b(f1) • Overconstrain the “Unitarity Triangle” Test the SM David Lange, LLNL

  3. fCP Three observable interference effects • CP violation in mixing • (direct) CP violation in decay • (indirect) CP violation in mixing and decay Add A= here David Lange, LLNL

  4. BB events are large fraction of the “physics” cross section (=1 nb) Coherent production of B meson pair (in L=1 state) Need high luminosity to produce sufficient event samples m(U(4S)) ~ 2*m(B) Take advantage of known B momentum in COM. Spend ~12% of running time below BB threshold to generate qq “continuum” events (ie, background samples for CP analyses. BB production threshold • Off • On CP Physics at the U(4S) BB enriched sample Fantastic Pep II performance allows us to study CP violation David Lange, LLNL

  5. SLAC B Factory performance • PEP-II top luminosity: 4.60 x 1033cm-2s-1 (exceeded design goal 3.0 x 1033) • Best 24 hours: 308.8 pb-1 Total Luminosity:: • PEP-II delivered 99 fb-1 • BaBar recorded 94 fb-1 Most analyses in this talk: On peak data 81 fb-1 # of BB pairs 88M Off peak data 10 fb-1 9 GeV e- on 3.1 GeV e+ Boost bg = 0.55 IP beam size 147 mm x 5 mm David Lange, LLNL

  6. BaBar Detector All subsystems crucial for CP analysis SVT: 97% efficiency, 15 mm z hit resolution (inner layers, perp. tracks) SVT+DCH: (pT)/pT= 0.13 %  pT+0.45 % DIRC: K- separation 4.2  @ 3.0 GeV/c  2.5  @ 4.0 GeV/c EMC:E/E = 2.3 %E-1/4 1.9 % David Lange, LLNL

  7. Quartz bar Active Detector Surface Particle Cherenkov light Detector of Internally Reflected Cherenkov Light (DIRC) • Measure angle of Cherenkov Cone in quartz • Transmitted by internal reflection • Detected by PMTs cos qc = 1/nbp=mbg David Lange, LLNL

  8. K/p Separation with the DIRC • Cherenkov angle c resolution and K- separation measured in data • Excellent K-p separation up to kinematic endpoint for B decay products. • Crucial for identification of charmless decays and for B flavor tagging. David Lange, LLNL

  9. no CPV Direct CPV: Interference of Decay Amplitudes Time-independent CP observable: Partial decay rate asymmetry 2 amplitudes A1 and A2 Strong phase difference Weak phase difference For neutral modes, direct CP violationcompetes with other types of CP violation • Large ACP requires amplitudes of similar order • bu: suppressed tree: charmless decays • large predicted ACP • bs: penguins: radiative decays • small predicted ACP • Understand penguins • Access to a and g • New Physics in loops David Lange, LLNL

  10. signal B background Event Selectionfor fully Reconstructed B mesons Use known beam (=EB) energy improve E and p conservation constraints DE s ~ 15-80 MeV; larger with neutrals mESs ~ 3 MeV David Lange, LLNL

  11. u,d,s,c background Signal MC Fisher Discriminant Overview of Charmless/Rare B Analyses • Analysis issues: • BR ~ 10-5-10-6 need lots of data • Large background from e+e- qq  background suppression • Modes with p0 suffer backgrounds from other B decays • Maximum likelihood (ML) fits to extract results • Kinematic and topological information separate signal from light-quark background • Particle ID to separate pions and kaons • Beware of charge bias • detector: trigger, tracking; reconstruction • Event selection, particle ID, analysis Arbitrary normalization David Lange, LLNL

  12. BK+p0 BK+p- BK-p0 BK-p+ Two body Charmless B decays Sensitive to:  eg. B+-, +0, 00  eg. BK+-, K0+ using Fleischer Mannel bound ACP can be sizeable see Paul Bloom’s talk Plots have an optimised cut on likelihood ratio David Lange, LLNL

  13. ~88106 B pairs New Results Preliminary Results ~60106 B pairs hep-ex/0207065 K+p-: hep-ex/0207055(Sub. to PRL) 23 21 hep-ex/0206053 5% ACP sensitivity in BK+p- David Lange, LLNL

  14. Direct CPV: B -D0(CP)K - D0K+K - DE • One ingredient to theoretically clean measurement of g. • Experimentally very difficult • [M. Gronau and D. Wyler, Phys. Lett B265, 172 (1991)]: hep-ex/0207087 Preliminary David Lange, LLNL

  15. Summary of (time integrated) Direct CP results • Details in: hep-ex/0207065, hep-ex/0206053, hep-ex/0207055, hep-ex/0207087, PRL88 101805, PRD65 091101, PRD65 051101. David Lange, LLNL

  16. CP Formalism for CP from Interference CP violation results from interference between decays with and without mixing mixing Amplitude ratio decay Time-dependent CP Observable: cosine term sine term (DG=0) David Lange, LLNL

  17. m- Flavor tag and vertex reconstruction K- Btag m- Brec m+ B0 KS B0 p+ Coherent L=1 state p- Start the Clock Stop the Clock Experimental technique at the (4S) resonance e+e-  (4S)  BB Boost: bg= 0.55 (4S) Exclusive B meson and vertex reconstruction David Lange, LLNL

  18. Tagging errors and finite Dt resolution dilute the CP asymmetry perfect tagging & time resolution typical mistagging & finite time resolution (f+) (f-) C=0 • Must determinemistag fraction wandDt resolution function Rin order to measure CP asymmetry. • Fundamental assumption: w and R are the same for CP events and more plentiful Brec modes. Measure from data with B0-B0 decays to flavor eigenstates. David Lange, LLNL

  19. Use self-tagged Bflav sample to measure w and R B0D(*)-p+/ r+/ a1+ Ntagged=23618 Purity=84% B0 J/yK*0(K+p-) Ntagged=1757 Purity=96% • High statistics, known decay time distribution: Bflav sample is x10 size of CP sample David Lange, LLNL

  20. BREC direction BREC Vertex BREC daughters Interaction Point Beam spot TAG Vertex z BTAG direction TAG tracks, V0s Vertex and Dt Reconstruction • Reconstruct Brec vertex fromcharged Brec daughters • Determine BTag vertex from • All charged tracks not in Brec • Constrain with Brecvertex, beam spot, and (4S) momentum • Remove high c2 tracks (to reject charm decays) • High efficiency: 95% • Average Dz resolution ~ 180 mm (dominated by BTag) (<|Dz|> ~ 260 mm) • Dt resolution function measured from data David Lange, LLNL

  21. Lepton tag: primary leptons from semileptonic decay Kaon1 tag: high quality kaons, correlated K+ andps- (from D*) Kaon2 tag: lower quality kaons, ps from D* Inclusive tag: unidentified leptons, low quality K, p, l No tag: event is not used for CP analysis B Flavor tagging method Exploit correlations between B flavor and its decay products to determine flavor of Btag. l - b c s K - Using tracks with or without particle identification, and kinematic variables, a multilevel neural network assigns each event to one of five mutually-exclusive categories: New and improved tagging method David Lange, LLNL

  22. Tagging performance from Bflav sample Measure of tagging performance Q: Q=e(1-2w)2 New tagging method increases Q by 7% compared to the method used in our previous result: PRL87 (Aug 01). David Lange, LLNL

  23. sin2b golden sample: (cc)KS and (cc)KL ccKs modes New • Data Signal J/y KL J/y Bkg Fake J/y Bkg Cos coeff. = 0 Sin coeff. = -hCPsin2b Theoretically “clean” and experimentally “easy” channels David Lange, LLNL

  24. Fit results hf =-1 hf =+1 sin2b = 0.755  0.074 sin2b = 0.723  0.158 sin2b = 0.741  0.067 (stat)  0.033 (sys) Preliminary David Lange, LLNL

  25. Golden modes with a lepton tag The best of the best! Ntagged = 220 Purity = 98% Mistag fraction 3.3% sDt 20% better than other tag categories background Consistent results across mode, data sample, tagging category sin2b = 0.79  0.11 David Lange, LLNL

  26. Sources of Systematic Error s(sin2b) Description of background events 0.017 CP content of background components Background shape uncertainties Composition and content of J/y KL background 0.015 Dt resolution and detector effects 0.017 Silicon detector alignment uncertainty Dt resolution model Mistag differences between BCP and Bflav samples 0.012 Fit bias correction 0.010 Fixed lifetime and oscillation frequency 0.005 TOTAL 0.033 Steadily reducing systematic error: July 2002 = 0.033 July 2001 = 0.05 David Lange, LLNL

  27. Search for non-Standard Model effects in (cc)KS • If another amplitude (new physics) contributes a different phase, then |lf|!=1 (or Cf !=0) • Fit |lf| and Sf using the (cc)Ksmodes (DG=0) | lf| = 0.948  0.051 (stat)  0.017 (syst) Sf = 0.759  0.074 (stat)  0.032 (syst) Consistent with the Standard Model expectation of |lf|=1 and nominal fit sin2b = 0.755  0.074 for (cc)Ks modes alone. David Lange, LLNL

  28. Other modes with ACP(Dt) proportional to sin2b • Compare with “golden” measurements to test consistency of CKM picture • Differences = Penguin “pollution” or New Physics David Lange, LLNL

  29. D* - Vcd B0 D* + D* - B0 D* + (b ccd) mode B0D*+D*- • Tree level weak phase same as b ccs • Penguin contribution unknown: • expected to be small (< 0.1*Tree) • Not a CP eigenstate, mixture of CP even (L=0,2) and CP odd (L=1) • Resolve using angular analysis Ntagged= 102 Purity = 82% David Lange, LLNL

  30. CP composition of B0D*+D*- • Measure CP odd fraction: R = 0.07  0.06 (stat) 0.03 (syst) • Almost pure CP even state. Take advantage of this in CP fit. David Lange, LLNL

  31. CP asymmetry fit • Improved fitting strategy: • Parameterize in terms of CP even (l+) and odd (l) components, include angular information from partial-wave analysis • Fix CP odd component to l=1, Im(l) = -0.741 • We measure: |l+|= 0.98  0.25 (stat)  0.09 (syst) Im(l+)= 0.31  0.43 (stat)  0.10 (syst) If penguins negligible: Im(l+) = - sin2b Preliminary hep-ex/0207072 Im(l+)measurement ~2.7s from BaBar sin2bin charmonium, assuming no penguins. David Lange, LLNL

  32. D+ D*+ D _ B0 CP conjugation strong phase D _ D+ D*+ B0 D*_ D*_ (b ccd) mode B0D*+D- • D*D not CP eigenstate  added complication • Brp is similar (a) • Strong phase contribution (and still have penguins) • Different (but related) decay time distributions • Measure both S and C coefficients David Lange, LLNL

  33. (b ccd) mode B0D*+D- D*D 56 fb-1 Ntagged= 85 Purity = 52% 56 fb-1 S+- = -0.43  1.41  0.20 C+- = 0.53  0.74  0.13 S-+ = 0.38  0.88  0.05 C-+ = 0.30  0.50  0.08 Preliminary Update to full data set in progress David Lange, LLNL

  34. (bccd) mode B0J/yp0 • Cabibbo and color-suppressed mode • Tree and penguin contributions could be comparable. Vcd Ntagged= 49 Purity = 59% hf = + 1 Tree: ~VcbVcd* ~ O(l3) same weak phase as bccs Penguin:~VcbVcd* + VubVud* ~ O(l3) adds additional weak phase David Lange, LLNL

  35. CP asymmetry fit for B0J/yp0 hf= + 1 background Preliminary In absence of penguins Cyp=0, Syp = -sin2b David Lange, LLNL

  36. sin2b from penguin mode B0fKS • Charmless decay dominated by (b sss) gluonic penguins • Weak phase same as b ccs. Sensitive to new physics in loops • Small branching fraction O(10-5) • Significant background from qq continuum • Using only fK+K- Ntagged = 66 Purity = 50% David Lange, LLNL

  37. Dt (ps) CP asymmetry fit for B0fKS • Low statistics. So: • Fix |lfK| = 1 • Fit for SfK background Preliminary +0.52 - 0.50 • SfK = -0.19 (stat)  0.09 (syst) If no new physics, SfK = sin2b David Lange, LLNL

  38. r = r (1-l2/2) h = h (1-l2/2) Standard Model comparison One solution for b is in excellent agreement with measurements of unitarity triangle apex Method as in Höcker et al, Eur.Phys.J.C21:225-259,2001 1 CKM angle down… David Lange, LLNL

  39. With Penguins (P): Bpp to measure sin2aeff No Penguins (Tree only): mixing decay Need branching fractions for p+p-, pp0, and p0p0 to get a from aeff isospin analysis David Lange, LLNL

  40. Time dependant analysis for Charmless B Decays 81/fb B h+h- sample split by tagging category • Analysis methods similar to yK with additional challenges • For example: The tagging efficiency is very different for signal and bkg • Strong bkg suppression in categories with the lowest mistag prob (Lepton/Kaon) Tagging Efficiencies (%) David Lange, LLNL

  41. Validation of Tagging, Vertexing, and ML Fit Fit projection in sample of Kp-selected events • Kp decays are self-tagging • T = tag charge • Q = kaon charge • Float t and Dmd in same sample used to extract CP asymmetries: David Lange, LLNL

  42. BppCP Asymmetry Results Fit projection in sample of pp-selected events Preliminary qq + Kp Submitted to Phys Rev (hep-ex/0207055) Using Grossman/Quinn bound (isospin only), combine with Bp+p0 and BABAR upper limit on Bp0p0: See P. Bloom talk for B->p0p0 David Lange, LLNL

  43. CP-Violating Asymmetries in B0 r+p-,r+K- R. Aleksan et al., Nucl. Phys. B361, 141 (1991) A. Snyder and H. Quinn, Phys Rev D48 2139 (1993) • Opportunity and challenges • In principle, can measure a directly, even with penguins • Much more difficult than p+p- • Three-body topology with neutral pion (combinatorics, lower efficiency) • Significant fraction of misreconstructed signal events and backgrounds from other B decays • Need much larger sample than currently available to extract a cleanly • We perform a “quasi-two-body” analysis: • Select the r-dominated region of the p+p-p0/K+p-p0Dalitz plane • Use multivariate techniques to suppress qq backgrounds • Simultaneous fit for r+p- and r+K- David Lange, LLNL

  44. Yields and Charge Asymmetries hep-ex/0207068 Preliminary David Lange, LLNL

  45. B0 rp time-dependent asymmetry Observables similar to D*D: hep-ex/0207068 Preliminary Systematic error dominated by uncertainty on B backgrounds David Lange, LLNL

  46. Searching for CP violating effects in time independent and time dependent studies of B meson decays. Growing # of direct CP violation searches sin2b from bccs (charmonium): (88M BB) Sin2aeff from Bpp :(88M BB) Much more than J/yK and pp: BD*D* B fK BJ/yp Brp More data required to turn these into “precision” measurements. Just completed long 20 month run. Machine and detector upgrades underway for improved luminosity performance. Conclusion and outlook sin2b = 0.741 ± 0.067 (stat) ± 0.033 (syst) David Lange, LLNL

More Related